Antigen specific T cell activation and proliferation are essential for significant cellular immune responses. Antigen-presenting cells (APCs) such as dendritic cells (DCs) activate T cells through a multi-signal mechanism. The first signal is antigen specific and causes T cells to enter the cell cycle. Other signals include costimulatory molecules that interact with ligands on the T-cell surface and leads to T-cell cytokine production and their proliferation. Mature DCs express several costimulatory molecules and are believed to be the most potent APCs. Clinical trials with mature DCs have demonstrated the induction of antigen specific T cell responses. A significant problem with DC based vaccines is determining the optimal method to uniformly mature the DC. A meritorious solution would be to genetically engineer DC to express the appropriate costimulatory molecules at high levels. Therefore, recombinant poxvirus vectors (replication-defective avipox [fowlpox; rF] have been engineered to express a TRIad of CO stimulatory Molecules (B7-1, intercellular adhesion molecule-1, and leukocyte function-associated antigen-3; designated TRICOM). The ability of dendritic cells to activate both naive and effector T cells in vitro and in vivo can be enhanced with the use of this poxvirus vector due to the hyperexpression of this triad of costimulatory molecules. In preclinical studies, murine DC infected with these vectors had an enhanced capacity to stimulate T-cell responses in allospecific mixed lymphocyte cultures, and to stimulate antigen specific CD8(+) T cells. The TRICOM vector worked most efficiently when used to infect murine DC ex vivo with DC, then administered as a cellular immunotherapy. To extend these promising observations to human studies, we have demonstrated that fowlpox engineered to express B7- 1, ICAM-1 and LFA-3 and the tumor antigen carcinoembryonic antigen (CEA) can infect human DC, express these molecules and stimulate CEA specific T cells in vitro. Therefore, we propose a phase I clinical trial to explore the safety, feasibility, and clinical and immunologic activity of this potent, novel and broadly applicable form of active immunotherapy- rF-CEA(6D)/TRICOM modified DC. We hypothesize that the use of DC engineered to hyperexpress costimulatory molecules will be a safe method to induce tumor antigen specific T cell responses in cancer patients, and these T cell responses will have clinical anti- tumor effects.